zoukankan      html  css  js  c++  java
  • python Hisogram

    # -*- coding: utf-8 -*-
    """
    Created on Fri Oct 24 19:32:45 2014
    
    @author: dell
    """
    
    # -*- coding: utf-8 -*-
    """
    Created on Fri Oct 17 09:14:25 2014
    generate figure of bootstrap test result.
    @author: dell
    """
    import numpy as np
    import matplotlib.pyplot as plt
    
    def myfun(array, myrange, mystep):
        countList = []
        for item in myrange:
            tp = filter(lambda x: item + mystep > x >= item, array)
            countList.append(len(tp))
        return countList
        
    if __name__ == '__main__':
        #egdf = pd.read_csv('fig_table.csv', index_col = 0)
        mydata = np.random.rand(100) * 100
        
        fig = plt.figure()
        # Ka
        ax = fig.add_subplot(311)
        myrange, mystep = np.linspace(mydata.min(), mydata.max(), 101, retstep = True)
        eglist = myfun(mydata, myrange, mystep)
        ax.plot(myrange, eglist)
        ax.set_title('Ka')
        ax.legend(['essential', 'non-essential'], 'best')
        # new Ka
        ax = fig.add_subplot(312)
        n, bins, patches = plt.hist(mydata, bins = 100, range = (mydata.min(), mydata.max()), histtype = 'barstacked')
        
        # bp histogram
        ax = fig.add_subplot(313)
        hist2, bins2 = np.histogram(mydata, bins = 100, range = (mydata.min(), mydata.max()))
        ax.plot(bins2[:-1], hist2)
        plt.show()

    Histogram

    自己随意试验的三种方法。

    注意的是 np.linspace(start, stop, num=50, endpoint=True, retstep=False) 指的是产生num个坐标,则整个line分为(num-1).

    自己写的子函数,用来算数量。

  • 相关阅读:
    A Bayesian Approach to Deep Neural Network Adaptation with Applications to Robust Automatic Speech Recognition
    nnet3的代码分析
    Kaldi中的L2正则化
    HMM拓扑与转移模型
    Kaldi阅读并更改代码
    nnet3中的数据类型
    nnet3配置中的“编译”
    Kaldi的delta特征
    Kaldi的交叉熵正则化
    【搜索】 Prime Path
  • 原文地址:https://www.cnblogs.com/hluo/p/4049133.html
Copyright © 2011-2022 走看看